change puppet server and clean certs on new rev 5 builder ref machine



7 years ago
5 months ago


(Reporter: jhford, Assigned: jhford)



we want to set the correct puppet server name on the machines coming up in scl3 and make sure the puppet client certs are removed so that the machines are able to sync with the scl3 puppet server when it comes up.

Amy, please comment with the exact fqdn of the scl3 puppet server.
Fwiw, until we have a working puppet server in scl3, there is no way to verify that things are working as expected.
Assignee: nobody → jhford
btw, the puppet master bug is bug 737590.
Jake, I've been toild that you'll be bring up the machine for me.  Please comment in this bug and ping me on IRC when the machine is powered on and ready for me to make the changes.  Please include the hostname of the machine
After a bunch of cleanup that was required on the client end, the host is up and responding and obtaining the correct hostname.

/Library/Preferences/SystemConfiguration/preferences.plist had information hard coded for the network and system configuration which needed to be deleted:

> 				<key>Addresses</key>
> 				<array>
> 					<string></string>
> 				</array>
> 				<key>Router</key>
> 				<string></string>
> 				<key>SubnetMasks</key>
> 				<array>
> 					<string></string>
> 				</array>
> 	<key>System</key>
> 	<dict>
> 		<key>Network</key>
> 		<dict>
> 			<key>HostNames</key>
> 			<dict>
> 				<key>LocalHostName</key>
> 				<string>noservers</string>
> 			</dict>
> 		</dict>
> 		<key>System</key>
> 		<dict>
> 			<key>ComputerName</key>
> 			<string>noservers</string>
> 			<key>ComputerNameEncoding</key>
> 			<integer>134217984</integer>
> 			<key>HostName</key>
> 			<string>noservers</string>
> 		</dict>
> 	</dict>
Oh, and to put this in the bug, the hostname is:
The edid emulator initially plugged into the second test machine (044) was defective.  After verifying the same display failures using the first machine (074) we were able to run puppet correctly without any human interaction.

Because scl3 is a new datacenter, I added a bogus master to slavealloc so that I could test that buildbot started correctly.  It did and buildbot is currently running on the second box trying to connect to this bogus master.

This image is good to deploy to the remaining 79 machines.  Please redo 074 as I am unsure of its state.
this image was deployed
Last Resolved: 7 years ago
Resolution: --- → FIXED
Product: → Release Engineering
Component: Platform Support → Buildduty
Product: Release Engineering → Infrastructure & Operations
You need to log in before you can comment on or make changes to this bug.